To investigate the concept of 'urethral atrophy', which is often cited as a cause of recurrent incontinence after initially successful implantation of an artificial urinary sphincter (AUS); and to investigate the specific cause of the malfunction of the AUS in these patients and address their management. Between January 2006 and May 2013, 50 consecutive patients (mean age 54.3 years) with recurrent incontinence had their AUS explored for malfunction and replaced with a new device with components of exactly the same size, unless there was a particular reason to use something different. Average time to replacement of the device was 10.1 years. The mean follow-up after replacement of the device was 24.7 months. All patients without an obvious cause for their recurrent incontinence had preoperative urodynamic evaluation, including measurement of the Valsalva leak point pressure (VLPP) and the retrograde cuff occlusion pressure (RCOP). After explantation of the AUS in patients without any apparent abnormality of the device at the time of replacement, the pressure generated by the explanted pressure-regulating balloon (PRB) was measured manometrically, when this was possible. In a select group of six consecutive patients of this type, the fibrous capsule surrounding the old cuff was incised then excised to expose and evaluate the underlying corpus spongiosum. In 31 of the 50 patients (62%) undergoing exploration, a specific cause for the malfunction of their AUS was defined. In the other 19 patients (38%) no cause was found, either preoperatively or at the time of exploration, other than a low VLPP and RCOP. A typical 'waisted' or 'hour-glass' appearance of the underlying corpus spongiosum was demonstrable, to some degree, on explanting the cuff in all cases. In the six patients in whom the restrictive sheath surrounding the cuff was excised, the urethral circumference immediately returned to normal after the compressive effect of the sheath was released. Manometry of the explanted PRBs, when this was possible, showed a loss of pressure in all instances. Replacement of the explanted AUS with a new device with the same size cuff and PRB in 14 of these 19 patients was successful in 12 (85.7%). These results, and other theoretical considerations, suggest that recurrent incontinence, years after initially successful implantation of an AUS, is because of material failure of the PRB, probably attributable to its age and consequent loss of its ability to generate the pressure it was designed to produce, and that urethral atrophy does not occur. Simply replacing the old device with a new one with the same characteristics, unless there is a particular reason to do otherwise, is usually successful and avoids the complications of alternatives such as as cuff downsizing, implanting a PRB with a higher pressure range, implantation of a second cuff or transcorporeal cuff placement, all of which have been advocated in these patients.